Aplicação da otimização robusta com conjunto incerto correlacionado no problema da designação de geradores com geração eólica

• Ano de defesa: 2017
• Autor(a) principal: Bombacini, Marcos Roberto
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal do Paraná; Curitiba; Brasil; Programa de Pós-Graduação em em Métodos Numéricos em Engenharia; UFPR
• Palavras-chave em Português: Programação (Matemática); Otimização matemática; Fator de potência; Programming (Mathematics); Mathematical optimization; Electric power factor; CNPQ::ENGENHARIAS::ENGENHARIA MECANICA; ENGENHARIA MECÂNICA
• Link de acesso: http://repositorio.utfpr.edu.br/jspui/handle/1/3429

The intermittency and stochasticity of the renewable resources bring significant challenges to the robustness and economic operation of power systems. Robust optimization allows for the modeling of an uncertainty set and ensures that the chosen solution can handle any possible realization based on this uncertainty set. Studies have shown that a large geographic spread of installed capacity can reduce wind power variability and the production more predictable. Current research work rarely considers the temporal and spatial correlation among different positions in wind farms. Motivated by these challenges, we present a robust optimization model for the unit commitment (UC) problem, which models the nature of the dispatch process and utilizes a new type of correlated uncertainty sets to capture the temporal and spatial corre-lations of wind farms. The results of applying the proposed model revealed that in the case of existing considerable correlation among wind power generation, it leads a superior solutions compared with that of conventional model by preventing protection against uncorrelated perturbations. IIn addition, the results revealed that the proposed model obtains superior performance as the values of correlations among the wind farms increases. As a conclusion, the proposed robust optimization model can be considered as an effective model for an environment containing correlated uncertain data.